A cross sectional epidemiological investigation was carried out of baggers, cashiers and produce clerks at two grocery stores in Oregon to detect possible cases of phytophotodermatitis after two cases were diagnosed in one store (store-A). Both stores were part of the same supermarket chain. The survey consisted of questionnaire interviews and dermatological examinations. Nineteen cases of phytophotodermatitis were detected, all occurring in the employees of store-A. Exposure to outdoor sunlight during a workshift and use of tanning salons were identified as risk factors. Severe (blistering) skin reactions were more frequent among employees who used tanning salons. Celery was the produce item most frequently handled by those with phytophotodermatitis. The cases handled celery 1.7 times more frequently than those with no dermatitis. Both stores carried the same brand of celery and sales volumes at both stores were usually comparable. Store-A had held a celery sale, however, coincident with the outbreak during which the amount of celery sold was about four times the usual volume. Photoprovocation testing using trimmed surfaces at the leafy ends of celery samples from store-A caused phototoxic reactions in three individuals. Chemical analyses of these samples showed elevated concentrations of psoralens (known phototoxic agents). The authors suggest that phytophotodermatitis in grocery store workers can be caused by healthy celery, and that it can result from a complex interaction of exposure variables such as ultraviolet radiation from tanning salons or natural sunlight, brand of celery and sporadic elevations of psoralen content of the celery.